Characterization and operation of a multi-channel Condensation Particle Counter (mc-CPC) for aircraft-based measurements

Richter, Sarah; Keber, Timo; Heinritzi, Martin; Beck, Lisa; Merkel, Laurin; Kirchhoff, Sarah; Schrod, Jann; Weber, Patrick; Curtius, Joachim

doi:10.5194/egusphere-2025-4349

Preprints

https://doi.org/10.5194/egusphere-2025-4349

Preprints

17 Sep 2025

| 17 Sep 2025

Characterization and operation of a multi-channel Condensation Particle Counter (mc-CPC) for aircraft-based measurements

Sarah Richter, Timo Keber, Martin Heinritzi, Lisa Beck, Laurin Merkel, Sarah Kirchhoff, Jann Schrod, Patrick Weber, and Joachim Curtius

Abstract. Field measurements of aerosol number concentration and aerosol size distribution in the upper troposphere and lowermost stratosphere (UTLS) are crucial for understanding the influence of processes such as new particle formation (NPF) on aerosol budgets, cloud formation and climate. In this study, we present the multi-channel Condensation Particle Counter (mc-CPC) that was designed and constructed for airborne measurements and tested during the TPEx campaign onboard a Learjet in 2024. The instrument uses FC-43 (C₁₂F₂₇N) as the working fluid and consists of three individual CPCs (Grimm SKY-CPC), a pressure regulation system and a common inlet. By varying the temperature difference (ΔT) between each pair of saturator and condenser, the individual cutoff diameters (d₅₀) of each CPC can be adjusted. For the cases presented here, we typically operated two of the CPCs at a ΔT of 36 °C for a direct comparison while the other CPC was set to a ΔT of 15 °C. Two independent calibration setups were used to determine the cutoff and size-dependent counting efficiency of the mc-CPC at various internal and external CPC pressure levels. The experiments in the laboratory showed that the cutoffs of the individual channels were rather independent of the external pressure p_external and only slightly dependent on the internal CPC pressure p_CPC, at least for a p_CPC range between 200–350 hPa. A large fraction of flights during TPEx were conducted at an internal pressure of 250 hPa, and therefore the cutoff determined at 250 hPa was used as a fixed value for all internal pressures. For channel 1 and 2 that were operated at the same ΔT, this gave a d₅₀ of 11.3 (±1.0) nm and 12.3 (±1.1) nm, respectively. Channel 3 was set to ΔT = 15°C and a cutoff diameter of 14.9 (±1.3) nm was determined. In an internal pressure range between 200 hPa and 400 hPa the cutoffs decreased slightly with increasing p_CPC. Furthermore, our measurements also indicate that the cutoffs are not influenced by varying sample flows. The mc-CPC was operated for the first time on an aircraft during the TPEx campaign (TropoPause composition gradients and mixing Experiment) in June 2024. We present the first measurements of one research flight and discuss the uncertainties of the collected aerosol data.

Received: 04 Sep 2025 – Discussion started: 17 Sep 2025

Competing interests: At least one of the (co-)authors is a member of the editorial board of Atmospheric Measurement Techniques.

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Sarah Richter, Timo Keber, Martin Heinritzi, Lisa Beck, Laurin Merkel, Sarah Kirchhoff, Jann Schrod, Patrick Weber, and Joachim Curtius

Status: closed

RC1:
'Comment on egusphere-2025-4349', Anonymous Referee #1, 09 Oct 2025
This paper presents a description of an aircraft-based, multi-channel condensation particle counter (mc-CPC) used for investigating the new particle formation (NPF) in the UTLS. They provide a detailed description of the system design and the careful, comprehensive calibrations, with an example of in-flight data from the TPEx campaign on tropopause composition in 2024. This mc-CPC system was developed by integrating three commercial GRIMM SKY-CPCs with a custom-built pressure regulation and flow manifold for aircraft-based measurements. FC-43 was used as the working fluid, which, according to the authors, was tested for the first time on Grimm SKY-CPC. The mc-CPC system consists of three channels to provide two size cuts, ~11-12 nm (chan1 and chan2) and ~15 nm (chan3). The counting efficiency of the CPC was corrected for flow and pressure, but not for the particle loss through the inlet and sampling line. The comparative use of the two size cuts from the mc-CPC (i.e., the difference between the particle number concentrations of the low and high size cuts) provides identification of NPF events.

Main comments:
Wording of “construct”. “We constructed a multi-channel condensation particle counter.” I am not sure if the wording of “construct” is entirely accurate here. It seems more like a custom integration built around commercial CPC units.

The biased particle concentration is a concern. The absolute particle number concentration from mc-CPC cannot reflect the true ambient values because the particle loss in the inlet and tubing is not corrected. The authors have attempted to estimate the particle loss during NPF events, but because the particle size distribution is unknown, the actual concentrations are still quite uncertain. The particle concentration of the two size cuts can be used comparatively for NPF event identification, because all CPCs share the same common inlet and sampling line. However, the loss of particles is heavily dependent on their size, especially for particles smaller than 20nm; as a result, the errors in absolute concentration between the two size cuts could bias the identification of NPF events.

The size cut for “recent NPF events.” The motivation of the mc-CPC is to investigate the NPF events in the UTLS, which, in this study, are identified by the difference in particle number concentrations between the lower size cut and higher size cut (12-15 nm). Can 12-15 nm (instead of sub-10 nm) be used to identify “recent NPF” (line 240)? Or maybe this is just another wording issue with the word “recent”.

The normalized counting efficiency and data correction. The normalized counting efficiency was shown in the manuscript to illustrate the instrument/channel comparison. However, the manuscript didn’t explicitly state whether the raw/absolute counting efficiency was used for in-flight data correction. There needs to be some statement to clarify this.

The Correction factor at 200 hpa (Pcpc). The internal pressure of the CPC (Ppcp) ranges from 200 hpa up to 750 hpa. However, the sample flow at 200 hpa is lower than that at higher pressures, particularly for FF>1.5 (line 743-744, Fig. B1). In the manuscript, the author stated that they cannot explain this behavior and did not account for these lower flows in the correction factor. However, the mc-CPC was operated at ~ 200 hpa for most of the time (i.e., RF04). If 200 hpa is a typical operating pressure for mc-CPC, and the outstanding flow behavior at 200 hpa is consistent, it needs further investigation and should not be simply ignored.

The manuscript could benefit from being more focused and concise, emphasizing the key points and minimizing redundancy.

Other comments
Need to keep consistency for the use of terms "NMP", "NMPs", and "nucleation mode particles". For example, line 56 uses both NMP and nucleation mode particles.

Line 79: a constant low pressure? According to the manuscript, the cpc pressure was regulated but not constant. Or does it set at a constant pressure for each flight? Please be clearer here.

Table 1: might need to list the constants used in the Antoine equation for Butanol and FC-43, and the corresponding references.

Line 164: “a low and constant pressure”, was the pressure here meant for P1 or P2? Need to clarify here.

Figure 1: Incomplete information, for example, the label of the IDP-3 pump is missing.

Mentioned the full name of the TPEx campaign multiple times in the manuscript (i.e., line 26, line 80, and line 170); only needs to mention the full name once to reduce redundancy.

Figure 2: For the cold reservoir of alpha pinene–shouldn’t the left tube insert deeper than the right tube?

Line 587-589: “We tentatively propose that altering diffusion rates in combination with the long mc-CPC inlet lines could have caused the dropping CPC performance with decreasing pressures.” Was it “decreasing pressure” or “increasing pressure”? Was the “dropping CPC performance” meant the deviated size cuts at increasing pressure for channel 3 (Fig. 8a)?

Line 730: a typo here: variied → varied

Line 744: The CPC flow is lower at 200 hpa, which is a typical operating pressure of the mc-CPC. However, here the outstanding behavior at 200 hpa was simply ignored. Has the sample flow at this pressure been measured more than once? Is this a consistent behavior? If so, I don't think this behavior can just be dismissed from the calculation of the correction factor.
Citation: https://doi.org/10.5194/egusphere-2025-4349-RC1
- AC1: 'Reply on RC1', Sarah Richter, 26 Nov 2025
  
  Please see the attachement for our response.
  
  Citation: https://doi.org/10.5194/egusphere-2025-4349-AC1
RC2:
'Comment on egusphere-2025-4349', Anonymous Referee #2, 16 Oct 2025

Excellent work in characterizing and correcting the performance of the CPC during non-ideal cases, such as unknown/uncontrolled sample flow and varying external and instrument pressures. There’s a lot of normalization that is happening that probably warrants a summation of error bars to understand the compounding corrections applied to the field dataset. The equation determining NPF is maybe overly strict, but also seems arbitrary as it's currently written. Considering this is the major focus of the instrument, I think this needs better justification or characterization. I think the combination of the aforementioned error analyses and Poisson variance would provide a stronger basis for quantifying NPF. Without the inlet pressure control being appropriately designed and characterized, the instrument may not accurately indicate the total particle number concentration, but is sufficient for NPF quantification since common-mode errors of larger particle loss is subtracted. This can be easily remedied in the next instrument update, to improve the amount of information received. Similarly, the second channel can be put to better use gathering additional information instead of redundant with the first channel.
L020: Why was flight 4 selected when it was operated at a different pressure from typical?
L089: Suggest changing “gets mixed” to “mixes” for succinctness.
L090: Suggest changing “enable” to “causing” or “activating”, showing causality.
L101: Critical orifice is not shown in Figure 1 flow diagram. Is it within the Channel block? What is the role of the pump valve if not for flow control?
L221: Correcting for inlet orifice particle loss isn’t necessary if only subtracting channels for an NPF measurement, but it is critical if trying to observe total particle concentration, especially in the lower Stratosphere, where mode size is typically accumulation mode. During ACCLIP, the integrated UHSAS concentration often exceeded the NMASS concentration in the Stratosphere due to losses in the pressure control orifice. Losses can be reduced with an appropriately designed expansion section after the orifice.
L241: Eqn. 2 values seem arbitrary. Some of the cited papers use different values for uncertainty. Recommend testing analysis from Williamson et al., 2019 with your dataset based on Poisson counting statistics.
L328: What was the DMA sample flow? Relevant to understanding transfer function width (and thus horizontal error bars in the sizing).
Fig 4: Can you clarify how the error bars are calculated? Is it the standard deviation from the varying particle number concentration measured during the time period of each test point? Is the standard deviation of the TSI reference CPC considered, since it will fluctuate as well. In Poisson statistics, need to sum variances together.
L472: This would not be a factor for sub-100 nm particles if an expansion section was included.
Fig 7: I think it would be nice to see the effect that pressure has on the plateau efficiency, rather than the normalized plots.
Table 4: Is Ch3 operating temperature a typo, or did you operate Ch3 at dT=36C? If so, why did you calibrate the counting efficiency outside of its operating spec? And why is D50 of CPC3 relatively high for the same dT?
Fig 9: How can you have N11-N15 ~= 1000 /scm3 and consider it “no NPF”? This circles back to my comments on Eqn. 2. The whole section from 12:30-13:45 looks strongly related to NPF.
L627: Is the inlet pressure reduction orifice properly sized for 200 hPa operation? Maybe you can go smaller and switch at a higher altitude to reduce amount of bypass flow the pump has to accommodate.
L641: The diffusion losses of 10-13 nm can be corrected since you have that information from subtracting the two channels and their relative contribution to the total concentration.
Instrument suggestions for future development: Designing and characterizing an appropriate expansion chamber at the inlet pressure control orifice will allow the first channel to (closer) measure a total number concentration. Now that the first and second channel have been tested together, I think the first channel alone can be trusted, so the second channel can be used for additional information such as a third cut size or a non-volatile measurement (if inlet orifice is characterized through Accumulation mode).

Citation: https://doi.org/10.5194/egusphere-2025-4349-RC2
- AC2: 'Reply on RC2', Sarah Richter, 26 Nov 2025
  
  Please see the attachement for our response.
  
  Citation: https://doi.org/10.5194/egusphere-2025-4349-AC2

Status: closed

RC1:
'Comment on egusphere-2025-4349', Anonymous Referee #1, 09 Oct 2025
This paper presents a description of an aircraft-based, multi-channel condensation particle counter (mc-CPC) used for investigating the new particle formation (NPF) in the UTLS. They provide a detailed description of the system design and the careful, comprehensive calibrations, with an example of in-flight data from the TPEx campaign on tropopause composition in 2024. This mc-CPC system was developed by integrating three commercial GRIMM SKY-CPCs with a custom-built pressure regulation and flow manifold for aircraft-based measurements. FC-43 was used as the working fluid, which, according to the authors, was tested for the first time on Grimm SKY-CPC. The mc-CPC system consists of three channels to provide two size cuts, ~11-12 nm (chan1 and chan2) and ~15 nm (chan3). The counting efficiency of the CPC was corrected for flow and pressure, but not for the particle loss through the inlet and sampling line. The comparative use of the two size cuts from the mc-CPC (i.e., the difference between the particle number concentrations of the low and high size cuts) provides identification of NPF events.

Main comments:
Wording of “construct”. “We constructed a multi-channel condensation particle counter.” I am not sure if the wording of “construct” is entirely accurate here. It seems more like a custom integration built around commercial CPC units.

The biased particle concentration is a concern. The absolute particle number concentration from mc-CPC cannot reflect the true ambient values because the particle loss in the inlet and tubing is not corrected. The authors have attempted to estimate the particle loss during NPF events, but because the particle size distribution is unknown, the actual concentrations are still quite uncertain. The particle concentration of the two size cuts can be used comparatively for NPF event identification, because all CPCs share the same common inlet and sampling line. However, the loss of particles is heavily dependent on their size, especially for particles smaller than 20nm; as a result, the errors in absolute concentration between the two size cuts could bias the identification of NPF events.

The size cut for “recent NPF events.” The motivation of the mc-CPC is to investigate the NPF events in the UTLS, which, in this study, are identified by the difference in particle number concentrations between the lower size cut and higher size cut (12-15 nm). Can 12-15 nm (instead of sub-10 nm) be used to identify “recent NPF” (line 240)? Or maybe this is just another wording issue with the word “recent”.

The normalized counting efficiency and data correction. The normalized counting efficiency was shown in the manuscript to illustrate the instrument/channel comparison. However, the manuscript didn’t explicitly state whether the raw/absolute counting efficiency was used for in-flight data correction. There needs to be some statement to clarify this.

The Correction factor at 200 hpa (Pcpc). The internal pressure of the CPC (Ppcp) ranges from 200 hpa up to 750 hpa. However, the sample flow at 200 hpa is lower than that at higher pressures, particularly for FF>1.5 (line 743-744, Fig. B1). In the manuscript, the author stated that they cannot explain this behavior and did not account for these lower flows in the correction factor. However, the mc-CPC was operated at ~ 200 hpa for most of the time (i.e., RF04). If 200 hpa is a typical operating pressure for mc-CPC, and the outstanding flow behavior at 200 hpa is consistent, it needs further investigation and should not be simply ignored.

The manuscript could benefit from being more focused and concise, emphasizing the key points and minimizing redundancy.

Other comments
Need to keep consistency for the use of terms "NMP", "NMPs", and "nucleation mode particles". For example, line 56 uses both NMP and nucleation mode particles.

Line 79: a constant low pressure? According to the manuscript, the cpc pressure was regulated but not constant. Or does it set at a constant pressure for each flight? Please be clearer here.

Table 1: might need to list the constants used in the Antoine equation for Butanol and FC-43, and the corresponding references.

Line 164: “a low and constant pressure”, was the pressure here meant for P1 or P2? Need to clarify here.

Figure 1: Incomplete information, for example, the label of the IDP-3 pump is missing.

Mentioned the full name of the TPEx campaign multiple times in the manuscript (i.e., line 26, line 80, and line 170); only needs to mention the full name once to reduce redundancy.

Figure 2: For the cold reservoir of alpha pinene–shouldn’t the left tube insert deeper than the right tube?

Line 587-589: “We tentatively propose that altering diffusion rates in combination with the long mc-CPC inlet lines could have caused the dropping CPC performance with decreasing pressures.” Was it “decreasing pressure” or “increasing pressure”? Was the “dropping CPC performance” meant the deviated size cuts at increasing pressure for channel 3 (Fig. 8a)?

Line 730: a typo here: variied → varied

Line 744: The CPC flow is lower at 200 hpa, which is a typical operating pressure of the mc-CPC. However, here the outstanding behavior at 200 hpa was simply ignored. Has the sample flow at this pressure been measured more than once? Is this a consistent behavior? If so, I don't think this behavior can just be dismissed from the calculation of the correction factor.
Citation: https://doi.org/10.5194/egusphere-2025-4349-RC1
- AC1: 'Reply on RC1', Sarah Richter, 26 Nov 2025
  
  Please see the attachement for our response.
  
  Citation: https://doi.org/10.5194/egusphere-2025-4349-AC1
RC2:
'Comment on egusphere-2025-4349', Anonymous Referee #2, 16 Oct 2025

Excellent work in characterizing and correcting the performance of the CPC during non-ideal cases, such as unknown/uncontrolled sample flow and varying external and instrument pressures. There’s a lot of normalization that is happening that probably warrants a summation of error bars to understand the compounding corrections applied to the field dataset. The equation determining NPF is maybe overly strict, but also seems arbitrary as it's currently written. Considering this is the major focus of the instrument, I think this needs better justification or characterization. I think the combination of the aforementioned error analyses and Poisson variance would provide a stronger basis for quantifying NPF. Without the inlet pressure control being appropriately designed and characterized, the instrument may not accurately indicate the total particle number concentration, but is sufficient for NPF quantification since common-mode errors of larger particle loss is subtracted. This can be easily remedied in the next instrument update, to improve the amount of information received. Similarly, the second channel can be put to better use gathering additional information instead of redundant with the first channel.
L020: Why was flight 4 selected when it was operated at a different pressure from typical?
L089: Suggest changing “gets mixed” to “mixes” for succinctness.
L090: Suggest changing “enable” to “causing” or “activating”, showing causality.
L101: Critical orifice is not shown in Figure 1 flow diagram. Is it within the Channel block? What is the role of the pump valve if not for flow control?
L221: Correcting for inlet orifice particle loss isn’t necessary if only subtracting channels for an NPF measurement, but it is critical if trying to observe total particle concentration, especially in the lower Stratosphere, where mode size is typically accumulation mode. During ACCLIP, the integrated UHSAS concentration often exceeded the NMASS concentration in the Stratosphere due to losses in the pressure control orifice. Losses can be reduced with an appropriately designed expansion section after the orifice.
L241: Eqn. 2 values seem arbitrary. Some of the cited papers use different values for uncertainty. Recommend testing analysis from Williamson et al., 2019 with your dataset based on Poisson counting statistics.
L328: What was the DMA sample flow? Relevant to understanding transfer function width (and thus horizontal error bars in the sizing).
Fig 4: Can you clarify how the error bars are calculated? Is it the standard deviation from the varying particle number concentration measured during the time period of each test point? Is the standard deviation of the TSI reference CPC considered, since it will fluctuate as well. In Poisson statistics, need to sum variances together.
L472: This would not be a factor for sub-100 nm particles if an expansion section was included.
Fig 7: I think it would be nice to see the effect that pressure has on the plateau efficiency, rather than the normalized plots.
Table 4: Is Ch3 operating temperature a typo, or did you operate Ch3 at dT=36C? If so, why did you calibrate the counting efficiency outside of its operating spec? And why is D50 of CPC3 relatively high for the same dT?
Fig 9: How can you have N11-N15 ~= 1000 /scm3 and consider it “no NPF”? This circles back to my comments on Eqn. 2. The whole section from 12:30-13:45 looks strongly related to NPF.
L627: Is the inlet pressure reduction orifice properly sized for 200 hPa operation? Maybe you can go smaller and switch at a higher altitude to reduce amount of bypass flow the pump has to accommodate.
L641: The diffusion losses of 10-13 nm can be corrected since you have that information from subtracting the two channels and their relative contribution to the total concentration.
Instrument suggestions for future development: Designing and characterizing an appropriate expansion chamber at the inlet pressure control orifice will allow the first channel to (closer) measure a total number concentration. Now that the first and second channel have been tested together, I think the first channel alone can be trusted, so the second channel can be used for additional information such as a third cut size or a non-volatile measurement (if inlet orifice is characterized through Accumulation mode).

Citation: https://doi.org/10.5194/egusphere-2025-4349-RC2
- AC2: 'Reply on RC2', Sarah Richter, 26 Nov 2025
  
  Please see the attachement for our response.
  
  Citation: https://doi.org/10.5194/egusphere-2025-4349-AC2

Sarah Richter, Timo Keber, Martin Heinritzi, Lisa Beck, Laurin Merkel, Sarah Kirchhoff, Jann Schrod, Patrick Weber, and Joachim Curtius

Viewed

Total article views: 1,425 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
1,285	103	37	1,425	25	28

HTML: 1,285
PDF: 103
XML: 37
Total: 1,425
BibTeX: 25
EndNote: 28

Views and downloads (calculated since 17 Sep 2025)

Month	HTML	PDF	XML	Total
Sep 2025	1,085	12	6	1,103
Oct 2025	108	31	8	147
Nov 2025	51	34	14	99
Dec 2025	41	26	9	76

Cumulative views and downloads (calculated since 17 Sep 2025)

Month	HTML	PDF	XML	Total
Sep 2025	1,085	12	6	1,103
Oct 2025	108	31	8	147
Nov 2025	51	34	14	99
Dec 2025	41	26	9	76

Viewed (geographical distribution)

Total article views: 1,327 (including HTML, PDF, and XML) Thereof 1,327 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 27 Dec 2025

Short summary

We constructed and characterized a Fluorinert multi-channel Condensation Particle Counter for aircraft applications that comprises three channels and a pressure regulation system to measure nucleation mode particles in the upper troposphere. The cutoffs were determined at several ambient and internal pressures and sample flows. During the first aircraft campaign TPEx we were able to identify possible new particle formation events.


Total:	0
HTML:	0
PDF:	0
XML:	0